2 Method
A random sample of each of the two populations was taken, with the aim of obtaining 250 valid responses in each case to a short survey document. A sample of 250 allows for confidence limits of about ± 6 percent at the 95 percent confidence level for an infinite population. Since both populations were unknown in size, but were bound to be relatively small, the anticipated confidence limits have been sustained.
The sample survey of travellers was collected at a series of stopping points within the lower Waitaki valley. These included the Maori rock drawings site at Takiroa, the Elephant Rocks Vanished World fossil site and the large fruit stall outside Kurow. Travellers were also encountered in Duntroon at the Vanished World information centre and the café and at various points within Kurow itself, such as the Visitor Centre and the Heritage Centre museum. Inevitably, this process made contact only with travellers who had stopped and is therefore biased against those who travelled straight through the district without stopping at all. A limited number of interviews were carried out outside the region with travellers who might have travelled through the area; these showed no significant difference from those undertaken in situ.
The sample of resident holidaymakers was aimed at people who were spending at least one night in the lower Waitaki valley and was drawn from the commercial accommodation sites identified in the previous study and from the fishing huts and similar properties at the river mouth. A small number of riverside freedom campers were also contacted. The support and cooperation of accommodation proprietors is gratefully acknowledged.
The survey forms used were designed for either interviewer completion or self-completion should that be necessary. Whenever respondents asked to take a survey form for later completion, they were provided with a postage paid envelope. In the event, virtually all of the traveller responses were collected by interview, whereas the majority of accommodation-based surveys was by self-completion.
Copies of each survey form are included in Appendix 1.
The surveys were carried out between 13 and 21 January 2005. The interview period occurred at a time when the high numbers associated with the Christmas and New Year period had departed and before a second peak in early February, associated with the commencement of the school year and the return of most families to their place of residence. As it transpired, numbers of both travellers and holidaymakers were relatively low and it was not possible to collect full samples with the time and resources available. In total, 201 valid responses were obtained from travellers through the area. Of those approached, very few declined to participate, but the survey staff noted that, at each survey point, many more vehicles passed by than stopped.
Rather fewer responses were obtained from the sample of those in accommodation in the area and the response rate for self-completion returns was also lower at slightly over 80 percent, although this is still very satisfactory for a self-completion survey. In particular, there were far fewer people in residence than was expected at the mouth of the river. Contacts there explained that the survey period was in the trough between the immediate Christmas period and the high levels of occupancy expected during the salmon run of late February to early April. Similarly, very few freedom campers were encountered, either. In all, a total of 124 useable responses were collected.
Responses to each question were coded numerically and the data files analysed through the use of SPSS, a standard social science statistical package.
